The term “monolithic structure” refers to mechanical elements that are inherently part of structures that receive process fluids. Typically, these monolith structures are unitary in design and can take the form of lids for fluid receiving chambers, such as chemical canisters (also known as chemical ampoules), semiconductor wafer processing chamber lids, single-gas and multiple-gas manifolds, and other pressure/vacuum chambers. Fluid communication into, and out of, these process fluid receiving chambers, as well as the control of the fluid, has been for many years managed by discrete fluidic assemblies typically consisting of valves, regulators, pressure transducers, and mass flow controllers, interconnected by reusable seals employing metal or elastomer gaskets and/or welding of tubular interconnects. In turn, these fluidic assemblies have communicated with the points of fluid use, or points of fluid origin, such as chemical canisters, wafer processing chambers, and gas manifolds, through reusable seals and/or welded attachments.
FIG. 1 show an example of a prior art external fluidic assemblies connected to a chemical canister. Typically, the primary inlet and outlet control valves are directly welded to the canister. In other prior art configurations, additional valves, and other components such as pressure transducers, may be part of the complex fluidic control arrangement permanently attached to the chemical canister. Several problems are caused as a result of these configurations. The valves, due to wear, contamination, or malfunction, can require replacement, the execution of which may be difficult, expensive, or simply impractical due to the permanent nature of the weld joints. Further, the use of weld-type valves and other tubulated fluidic components significantly increases the overall space required for installation and use of the canister. Further still, the volume and surface area of the externally connected components represent increased volume that must be evacuated or purged clean prior to filling of the canister or use of the contents following installation at the point of use. Further, and not insignificantly, the welded valves and associated fluidic components are all too often used as handling and carrying features, a practice that can lead to weakening and failure of the welded joints between the components and at the canister joints.
The advent of modular surface mount (“MSM”) fluidic components, beginning in the mid-nineteen nineties, is perceived as a significant milestone in reducing the size of fluidic systems. That is, systems comprised of fluid control and measurement components such as valves, regulators, filters, pressure transducers, mass flow meters, and mass flow controllers. Prior to MSM interfaces, such components were typically joined for fluid communication by interconnecting tabulations either via welding or via reusable gasketed connections. Either method was enabled by metal tubing protrusions, or appendages, intrinsic to each fluidic component for the express purposes of interconnection and fluid transport.
MSM interfaces did reduce the size, or “footprint,” of fluidic systems considerably. In MSM architecture, the fluidic component is sealed, typically with elastomer O-rings or metal gaskets, using bolts for compression, to a receptive MSM or “modular” architecture. Several MSM or modular architectures that are in common use are described in U.S. Pat. No. 5,836,355; U.S. Pat. No. 6,951,226; and U.S. Pat. No. 7,048,008. A common aspect of these disclosures is twofold: (1) to provide for the standardized fluidic interface to seal to the MSM component and (2) to provide interconnecting gas conduits for the purpose of routing fluids into, out of, and between fluidic components.
The reduction of size and internal “wetted” area and volume afforded by modular fluidic systems are well understood, especially within the semiconductor wafer processing industry wherein size, purity of fluids, cleanliness of the gas system, and serviceability are prized attributes of any fluidic system.
Although MSM-type fluidic systems offer advantages in terms of reduced size, reduced area and volume exposed to the controlled fluids, and improved serviceability, the MSM component typically must be sealed to a receptive modular architecture in the manner disclosed by the aforementioned patents. Put in another way, the MSM fluid component is typically mated to a corresponding modular interface in order to complete the fluidic circuit. Conventionally, this corresponding modular interface is provided by modular architectures of various designs but all of which embody the standard modular interface as set forth by SEMI Standards F86-0304 and F87-0304, among others.
There are, however, some exceptions to the convention of mating MSM fluidic components to standard MSM architectures. For example, it is possible to provide the appropriate mating interface on a non-modular, or monolithic, surface, but this requires a method of fluid communication to, or from, or between, the fluidic components. This may be accomplished by boring interconnecting fluidic passages in the monolithic structure itself, typically a stainless steel alloy. FIG. 2 shows such a configuration, with two MSM components, such as valves, regulators, pressure transducers, or filters, mounted to a monolithic base that has been machined with the sealing and bolt mounting configuration consistent with current MSM interface practice. Sealing of the MSM component directly to the surface of the monolithic structure is accomplished by using appropriate O-rings or metal gaskets. Additionally, interconnecting fluidic paths are machined into the monolithic structure at angles appropriate for the fluidic interconnection, or intercommunication, of the fluidic components. Additional fluidic paths may be machined to allow external fluid communication with the components, and, through them, to the process or use served by the monolithic structure. Referring again to FIG. 2 the process fluid is available at external port, P1, and can flow to or from fluidic component FC1, and from FC1 flow to or from fluidic component FC2, and from fluidic component FC2 to or from the chamber, to which monolithic structure MS1 is affixed, through port P2. The direction of flow through the fluidic circuit being determined by the pressure differentials present between P1 and P2. The monolithic structure, MS1, may be an integral part of a chemical canister used for the supply of chemical, or MS1 may be a structure integral to a vacuum or pressure processing chamber such as a semiconductor wafer processing chamber. Further, MS1 may be integral to yet another fluidic conduit for the purpose of controlling fluid into or out of that conduit.
While the general method for employing MSM components as depicted in FIG. 2 is known to the art, the method has several problems. The first problem is simply the cost of machining angled fluidic passages in materials such as stainless steel, nickel, titanium, Hastelloy, and other metals generally valued, and used, for their corrosion resistance in chemical and vacuum service. In applications involving ultra-high purity chemical delivery and/or ultra-high vacuum, it is desirable to smoothly converge, or blend, the angled fluidic passages where they are joined to produce very smooth surface finishes minimizing molecular entrapment, corrosion sites, and outgassing. These principles are well understood by those skilled in the art of ultra-pure chemical delivery or ultra-high vacuum. The need for near-perfect blending of the angled and convergent fluid paths results in demanding and often very expensive machining operations. A second problem associated with angled fluidic passages is one of reduced conductance. As anyone skilled in the subject of plane geometry may appreciate, a cylinder, when intersecting a plane at an angle, will produce an ellipse at the plane of intersection. And so it is with fluidic passages angled into monolithic structures and the resulting ovality of the cross-section of the passages at the plane of sealing. The standard diameter of fluidic passages in MSM components and corresponding modular architectures is 0.180″ diameter, established as such to emulate the inside diameter of commercial ¼″ stainless steel tubing that has been used predominantly for decades in liquid chemical and gas transport, and analytic instrumentation.
When a 0.180 diameter fluidic passage is machined into a monolithic structure at an angle of 30 degrees, an angle that may be considered typical in monolithic structure fluid passages, the an oval cross-section results from the intersection of the gas passage with the surface of the monolithic structure. It is on this surface, and coaxial with the fluidic passage, that the sealing feature for the O-ring or metal gasket, necessary to seal the MSM component to the passage, must be provided. The MSM standards for interface seals requires that the fluidic passage be no larger than 0.180″ diameter at the risk of compromising the O-ring seal or metal gasket seal. However, the major axis of the ellipse produced by machining the 0.180″ diameter gas path at 30 degrees from perpendicular is 0.208″ and would result in a compromised MSM seal. Thus, as is the present practice, the diameter of the internal fluidic gas passage must be reduced according to the angle of penetration so as to maintain the major axis of the resulting ellipse within the 0.180″ maximum dimensional standard. In practice, it is not uncommon to find gas passages, machined at acute angles, as small as 0.090″ diameter to avoid exceeding the 0.180″ diameter at the MSM sealing surface. The effect of this practice is dramatically reduced fluid conductance that may compromise the performance of the fluidic system. This is an especially important problem in fluidic systems controlling high viscosity liquids or low vapor pressure gases.
A possible solution to maintaining acceptable elliptical dimensions at the sealing surface for internally machined fluid passages is to manufacture them at angles closer to perpendicular to the sealing surface, and deeper. This is another simple matter of trigonometry wherein the termination of intersecting fluid passages in the lateral direction may be achieved by depth rather than angle. For example, one pair of passages could be constructed at a 30 degree angle from perpendicular while another pair is constructed at 10 degrees from perpendicular. Both solutions are designed to maintain a major elliptical axis of 0.180″ at the sealing surface. The result is that fluid passages fabricated at 30 degrees must be 0.156″ diameter, and the fluid passages fabricated at 10 degrees must be 0.177″ diameter in order to follow the guidelines for MSM sealing surfaces. As important, the fluidic passages fabricated more closely to perpendicular must be considerably deeper to achieve blended convergence. This reveals the basic design tradeoff required when attempting to use MSM components directly on the surface of monolithic structures: a tradeoff between conductances of the fluidic passages versus the depth required for their fabrication. In all cases, as explained earlier, the complexity and cost of machining angled interconnecting fluidic passages in monolithic structures is not inconsiderable. Further complicating this matrix of design decisions is the reality that few monolithic structures have sufficient thickness to permit near-perpendicular angles for the interconnecting fluidic passages and the resulting depths of the fluidic passages. Therefore it has become customary to sacrifice fluid conductance by decreasing the diameters of the fluidic passages and fabricating the passages at angles typically between 30 degrees and 45 degrees.
What is needed is a system for the direct implementation of MSM components on the surfaces of monolithic structures that allows maximum conductance consistent with the diameters of the fluidic passages of the components themselves. Further, this method must minimize the depth of required penetrations into the monolithic structures and eliminate the need and expense of fabricating angled fluidic passages. My invention addresses the various difficulties associated with the use of angled fluidic passages integral to monolithic structures, as described above, by using a small family of fluidic inserts embedded in a slot formed in the monolithic structure. The use of these inserts requires only the relatively simple machining of slots into the surfaces of monolithic structures for placement of the inserts and the subsequent sealing of the MSM fluidic components to the inserts by the fastening of the components to the surface of the monolithic structure. These and other features of my invention are described below.